trade openness, financial depth and economic … · 1 trade openness, financial depth and economic...
TRANSCRIPT
1
Trade openness, financial depth and economic
development: Causality evidence from Asia-Pacific
countries
Thai-Ha Le*
Department of Economics, Finance and Marketing
Centre of Commerce and Management, RMIT University, Vietnam
702 Nguyen Van Linh Blvd., District 7, HCMC, Vietnam
Email: [email protected]; Telephone: +84-983502188
* Corresponding author.
2
Trade openness, financial depth and economic development:
Causality evidence from Asia-Pacific countries
ABSTRACT
This study investigates the causality patterns between financial deepening, trade openness
and economic development for 14 countries in Asia and the Pacific. The Gregory Hansen
cointegration tests which account for one endogenous structural break and Toda-Yamamoto
non-Granger causality tests are used to add to the existing empirical evidence. In general, the
evidence indicates (1) a strong link between financial depth and economic development, (2) a
somewhat weaker linkage between financial depth and trade openness and (3) a sceptical
linkage between trade openness and economic development, for most of the sample.
Key words: financial depth, economic development, trade openness, Toda-Yamamoto non
Granger causality test, Gregory Hansen cointegration test, Asia and the Pacific.
JEL: N2, O1, O43.
3
1. INTRODUCTION
The last decades have witnessed development strategies adopted by many economies that
prioritize the modernization of their financial systems. The countries of Asia and the Pacific
(henceforth AP) are not exceptional cases. The current remarkable growth in the AP financial
markets, particularly with the recent developments in credit markets, is certain to continue.
Efforts to develop financial markets are theoretically needed to foster critical economic
activities such as the capital allocation process, monetary policy implementation and
government borrowing. The current global economic situation further underscores the
compelling rationale for the development of sound and integrated financial markets in the
region. However, the effectiveness of such policies requires a causal relationship between
financial and real sectors.
Even though the relationships between trade, financial depth and economic development have
been extensively explored in literature, the majority of the studies have used a bi-variate
framework to examine the causal relationship between trade and economic development and
between financial and economic development (e.g., Shahbaz, 2012; Calderon and Lin, 2003).
However, it has been clear that the results obtained by conducting bi-variate causality test
might be invalid due to the omission of an important variable which affect both the variables
included in the causality model. As such, the introduction of a third variable in the causality
framework may not only alter the direction of causality but also the magnitude of the
estimates (Loizides and Vamvoukas, 2005).
Further, several studies have employed methods for cross-sectional data analysis with a hope
that the causalities between the variables of interest could be generalized (e.g., Yanikkaya,
2003; Harrison, 1996). Yet, the problem of using a cross-sectional method is that by grouping
countries at different stages of trade openness, financial and economic development, the
method could not take into account the country-specific effects of trade openness and
financial depth on economic development and vice versa. Particularly, it fails to explicitly
address the potential biases arising from the existence of cross-country heterogeneity, which
may lead to inconsistent and misleading estimates (Ghirmay, 2004; Casselli et al., 1996). To
avoid this backdrop, this study attempts to investigate the causalities among trade openness,
financial depth and economic development in a number of AP economies using a tri-variate
framework.
4
This research assesses whether financial depth has led to economic development in a sample
of AP countries as these markets are expected to play a further critical role in the world
capital markets for investment and risk management. The study investigates whether a policy
focus on financial sector development is appropriate for fostering development. Thus
causality between finance and economic development is tested, capturing indirect linkages
also by scrutinizing the relationship between financial depth and trade openness. This study
contributes to the existing literature by (1) using advanced econometric methods that are less
prone to the misspecifications that occur when testing for cointegration and causality, (2)
employing a composite finance indicator in order to proxy financial development in a broad
sense, and (3) taking into account the linkages between financial depth and trade openness
that allow for further impacts on economic development.
The balance of this paper is structured as follows. Section 2 reviews the related academic
literature. Section 3 describes the data, variables and the testing framework. Section 4
presents and discusses the empirical results. Section 5 concludes with a summary.
2. BACKGROUND AND OVERVIEW
(a) Literature review
(i) Trade openness and economic development
Conventional trade theory proposes that international trade is associated with a reallocation of
resources within the national borders determined by exogenous differences across countries.
This reallocation of resources generates efficiency gains that lead to an increase in the level
of aggregate national income. Krugman (1979, 1980) claims two other sources of gain from
international trade. First, there could be more varieties of products available for consumption.
Second, the increased competition lowers the market power of firms and hence the
equilibrium prices. The lower prices raise real purchasing powers, which is another source of
gain for consumers. Further, the increased size of the market allows firms to realize
economies of scale. Even though the size and distribution of the welfare gains from trade may
be disputed, there is strong consensus of a positive relationship between international trade
and aggregate national income.
5
The same degree of consensus, however, does not appear to hold for the growth effects of
international trade. New growth theories do not predict that trade will unambiguously raise
economic growth. It is argued that increased competition could discourage innovation by
lowering expected profits (Schumpeter, 1934; 1942). Many empirical analyses estimate
positive growth effects of trade openness, but the size of these effects is often rather small.
However, it is argued that intervention in trade could raise long-run growth if protection
encourages investment in research-intensive sectors for countries with an international
advantage in these kinds of goods. Some empirical evidence suggests that trade openness
may indeed positively affect economic performance (e.g., Edwards, 1998; Harrison, 1996).
Since the theoretical literature shows a mixed answer, empirical work is needed to help
resolve the debate.
(ii) Financial depth and economic development
Financial markets, at a very broad level, are the venues where borrowers and lenders interact,
and capital is raised for real investment and then gets reallocated among investors. Liquid and
deep financial markets sway economic development. Financial development contributes to
increased mobilisation of savings as well as a reduction in information asymmetries, which
leads to better allocation of resources. Further, developing liquid financial markets is
essential for governments and central banks for the conduct of their fiscal and monetary
policy implementation. At a micro level, financial development involves improved
monitoring of managers and a higher level of corporate control which facilitates risk
reduction (King and Levine, 1993). A number of theoretical models have been proposed to
analyse the linkage between financial depth and economic growth (e.g., Levine, 2005).
The debate regarding the direction of causality between financial development and economic
growth has been ongoing since the 19th
century. The first view argues that financial
development leads to economic growth due to its influence through the accumulative and the
allocative channel. The accumulative channel emphasizes the finance-induced effects of
physical and human capital accumulation on economic growth (e.g., Pagano, 1993).
Meanwhile, the allocative channel focuses on the finance-induced gains in resource allocation
efficiency which translates into augmented growth (e.g., King and Levine, 1993). The second
view maintains that economic growth drives the development of the financial sector. For
instance, in an expanding economy, the private sector may demand new financial instruments
6
and a better access to external finance. As such, the finance activities simply amplify instep
with general economic development (e.g., Robinson, 1952). The third view contends that
finance and growth may be mutually dependent. The real sector may provide the financial
system with the funds necessary to enable financial deepening, eventually allowing for a
capitalization on financial economies of scale which in turn facilitates economic development
(e.g., Berthelemy and Varoudakis, 1996). Finally, the fourth view follows more sceptical
views that finance and growth may also evolve independently of each other, so no causality
(or insignificant causation) exists between them (Chandavarkar, 1992).
The majority of empirical studies on the relationship between finance and growth are cross-
sectional studies based on cross-sectional regressions. They documented a positive
connection between financial development and economic activity (e.g., King and Levin,
1993; La Porta et al, 2002). None of these cross-country studies, however, gave a satisfactory
answer to the causality question between financial depth and economic growth. Compared
with cross-country studies, in studies of individual countries, researchers can design specific
measures of financial development according to the particular characteristics of the country.
These studies can also avoid dealing with country-specific factors in regression analysis.
(iii) Financial depth and trade openness
It is shown that the countries with a relatively well-developed financial sector have a
comparative advantage in industries and sectors that rely on external finance (Kletzer and
Bardhan, 1987). Extending this argument and allowing both sectors to use external finance,
one being more credit intensive due to increasing returns to scale, the level of financial
development is found to have an effect on the structure of the trade balance (Beck, 2002). On
the one hand, reforming the financial sector might have implications for the trade balance if
the level of financial development is a determinant of countries’ comparative advantage. On
the other hand, the effect of trade reforms on the level and structure of the trade balance
might depend on the level of financial development. More recently, in building a model with
two sectors, one of which is financially extensive, Do and Levchenko (2004) find that
openness to trade will affect demand for external finance, and thus financial depth, in the
trading countries. In particular, their model predicts that in wealthy countries, trade should be
related with faster financial development. On the contrary, in poor countries, more trade
7
should slow financial development, because these countries import financially intensive
goods rather than develop their own financial system.
(iv) Financial depth-trade openness links and economic development
Multi-causal linkages among trade openness, economic development and financial deepening
emerge from the evidence that not only financial development favourably impacts but the
extent of financial activity itself depends positively on growth (e.g., Bencivenga and Smith,
1998). This is because the cost of financial services carries a fix component that falls with the
volume of financial transactions. As such, financial markets will develop only when a
threshold level of income is attained. But, if financial outcomes are endogenous to economic
development, the question of interest would be how greater trade integration affects the state
of financial development itself.
Gries et al. (2009) contends that linkages between financial depth and trade openness could
allow for more complex paths to economic development. In particular, if increasing trade
openness contributes to a higher level of financial development, this may promote economic
growth where financial depth is found to enhance growth via the allocative and accumulative
channels. But if financial deepening induces trade openness, it may subsequently foster
economic development where openness to trade is found to be a growth factor.
Blackburn and Hung (1998) employs the well-known endogenous growth model of Romer
(1990) to explore the multi-causal relationships among trade openness, economic growth and
financial development. In the model, economic growth is driven by horizontal innovation in
intermediate goods, which are encouraged by expanding the markets for new goods, e.g.,
through trade liberalisation. This implies that more firms would enter the research sector and
seek for external financing of risky and independent research projects. This helps financial
intermediaries to better diversify their portfolios and decreases their default probability. As a
result, the agency cost related to the need for depositors to monitor the intermediary portfolio
is reduced. The reduction in the agency cost of financial intermediation leads to higher
economic growth. This is because firms in the research sector start operating at positive
profits and this encourages new firms to enter the market. The rate at which new process are
invented is thus increased. This is an indirect financial market’s gain from trade. Specifically,
trade liberalisation can accelerate innovations and the development of financial markets
8
through scale effects. Hence, there theoretically exists a complementary relationship between
trade and financial development.
(b) Economic development in Asia and the Pacific
Over the last 20 years, the AP region has continued to keep high economic growth rates
exceeding those in other regions. Having accounted for more than half of global economic
growth, the region has consequently come to be known as the growth centre of the global
economy. Furthermore, the scale is expanding. The overall Asia-Pacific economy is growing
faster than any other regional economy. It is anticipated to be larger than that of Western
Europe and be equal to that of the Americas (North and South) by 2025. However, the
countries (and territories) of the region are at various levels of economic growth. While
Australia, Japan, Republic of Korea, New Zealand, and Singapore are regarded as highly
industrialized countries, Bangladesh, Cambodia, China, India, Pakistan, and Vietnam are
categorized as low-income countries. Indonesia and Philippine could be regarded as middle
income countries, and Thailand and Malaysia be as high income countries.1
The economic potential of this region implies opportunities for robust financial systems to
develop in the region. However, the linkage between the development of the financial
markets and their role in economic development needed to be carefully considered. For
instance, it might need to evaluate how big a financial system should be to remain anchored
in real economic activity. Further, it is important to see if financial systems are being built to
serve economies, or economies are being made subservient to the needs of financial system.
The ambiguity of the empirical literature, based on the above discussion, provides an
additional motivation for this study.
3. DATA AND METHODOLOGY
(a) Variables and data
Annual time-series observations are used as they are sufficient to ensure the quality of the
analysis, as argued by Hakkio and Rush (1991). The choice of the countries to be included in
the sample of this study is due to the availability of comprehensive data set. As for economic
1 Please refer to Chapter 1-1, ‘A Long-term Perspective on Environment and Development in the Asia-Pacific
Region’, available at: http://www.env.go.jp/en/earth/ecoasia/workshop/bluebook/chapter1-1.html
9
development, the logarithm of real GDP per capita (log-level data) is used and labelled as
OUTPUT. For trade openness, the logarithm of the sum of exports plus imports to real GDP
(log-level data) is used and labelled as OPENNESS because this measure is a simple and
common indicator of trade openness as suggested by Harrison (1996). The data are taken
from International Financial Statistics (IFS). As to financial development, there is a large
literature discussing its possible measures. In the related literature, several proxies for
financial depth have been suggested, for instance, money aggregates such as M2 to GDP
(e.g., Odhiambo, 2008) but there has been no consensus on the superiority of any indicator.
For measuring overall financial development, the most popular measure is the ratio of liquid
liabilities to GDP (LLGDP). Based on the liquid liabilities of the financial system, this
measure has been used in King and Levine (1993). This measure, however, can be too high in
countries with undeveloped financial markets. Other standard measures are the ratio to GDP
of credit issued to the private sector by banks and other financial intermediaries
(PCRDBOFGDP) and the ratio of the commercial bank assets to the sum of commercial bank
assets and central bank assets (DBACBA).
This study follows a recent method by Ang and McKibbin (2007) to construct a composite
indicator of financial deepening which is as broad as possible. Specifically the finance
proxies including LLGDP, PCRDBOFGDP and DBACBA are used to construct this index
labelled DEPTH via a principal component analysis. Since most financial systems in Asia are
bank-based, the financial indicators that are primarily associated with bank development are
used. Data for the individual finance indicators is taken from the updated and expanded
version of Financial Development and Structure Database (FDSD). The principal component
analysis reduces data sets to lower dimensions while retaining as much information of the
original sets as possible. In this case, the finance indicators are transformed into natural
logarithms and only the first unrotated principal component is extracted as DEPTH.
(b) Methodology
This study uses unit root and cointegration tests to identify the stationary properties and
possible cointegration relationships of the investigated time series. Specifically, the unit root
test by Phillips and Perron (1988), the PP test, is employed to check whether the considered
time series is stationary, that is, I(0), or first difference stationary, that is, I(1). The PP test is
used as it is particularly powerful when low frequency data are used (Choi and Chung, 1995).
10
Different methodological alternatives have been proposed in econometric literature to
empirically analyse cointegrating relationships between time-series variables (e.g., Engle and
Granger, 1987; Johansen, 1988; Johansen and Juselius, 1990). The cointegration frameworks
in these studies, however, have limitations when dealing with data as major economic events
may affect the data generating process. The presence of structural breaks in turn leads to
inefficient estimation and lower testing power (Gregory et al., 1996). The sensitivity of the
outcome of the tests to structural breaks has been documented in several studies (e.g., Lau
and Baharumshah, 2003). This study thus employed the Gregory and Hansen (GH hereafter)
(1996) tests for cointegration to analyse the long-run relationships and dynamics interactions
between time-series variables. The advantage of this test is the ability to account for the
possible presence of an endogenous structural break.
Following GH test, the TY methodology is employed to conduct causality test. Even though
the most common way to test for causal relationships between two variables is the Granger
(1969)’s causality test, it has probable shortcomings of specification bias and spurious
regression (Gujarati, 1995). The TY procedure improves the power of the Granger-causality
test. The procedure is a methodology of statistical inference, which makes parameter
estimation valid even when the VAR system is not co-integrated. This technique is applicable
irrespective of the integration and cointegration properties of the system, and fitting a
standard VAR in the levels of the variables rather than first differences like the case with the
Granger causality test. Thereby, the risks associated with possibly wrongly identifying the
orders of integration of the series, or the presence of cointegration are minimized and so is the
distortion of the tests’ sizes as a result of pre-testing (Mavrotas and Kelly, 2001). The method
involves using a Modified Wald statistic for testing the significance of the parameters of a
VAR(p) model where p is the optimal lag length in the system. The estimation of a
VAR(p+ guarantees the asymptotic distribution of the Wald statistic, where is
the maximum order of integration in the model. In this study, the lag lengths in the causal
models were selected based on the SC and the VAR was made sure to be well-specified by,
for instance, ensuring that there is no serial correlation in the residuals. If need be, the lag
length was increased until any autocorrelation issues are resolved. Needless to say, the
system must satisfy the stability conditions and the common assumptions to yield valid
inferences. The null of ‘no Granger causality’ is rejected if the test statistic is statistically
significant. That is, a rejection supports the presence of Granger causality.
11
4. EMPIRICAL RESULTS AND DISCUSSION
First, the principal component analysis is performed. Table 1 gives an overview of the results
of the principal component analysis and a descriptive overview of the investigated countries.
The index DEPTH used in this study is usually the only component to show fitting
characteristics. In all the cases, this index exhibits at least 60% of the initial variance of the
considered series and an eigenvalue that is significantly larger than one. Thus, the first
principal component captures adequately the three components of the DEPTH index.
[Please insert Table 1 here]
Next, this study uses the Phillips and Perron (PP)’s (1988) unit root test to check whether the
considered time series is I(0), that is, stationary, or I(1), that is, first difference-stationary.
The PP test is used as it is particularly powerful when the low frequency data are used (Choi
and Chung, 1995). As reported in Table 2, in almost all cases the PP test fails to reject the
null hypothesis of the existence of a unit root for the data at log level. Meanwhile, in all but
two cases the null hypothesis is rejected strongly when the first difference is taken. The
examined time series are thus I(1) at log level and I(0) at first log difference.
[Please insert Table 2 here]
Next, this study tests for cointegration in trivariate VAR models using log-level data,
following Gregory-Hansen (1996). Table 3a, 3b and 3c report the cointegration results for
trivariate VAR models using all three statistics: ADF*, and , with DEPTH, OUTPUT
and OPENNESS as the dependent variables in cointegrating equations, respectively. For
Korea, the results indicate two cointegrating relations between the series at 10% significance
level. For Japan, Nepal, China and Israel, the common suggestion is at most one cointegration
relationship at 10% significance level. When a cointegrating relationship is present, financial
depth, economic development and trade openness share a common trend and long-run
equilibrium as suggested theoretically. As to other countries in the sample of this study,
however, there is not enough evidence to conclude on the existence of cointegration between
the three series.
[Please insert Table 3a, b, c here]
The unit root test results indicate that the maximum order of integration among the variables
of interest is 1. Based on this, this study performs Toda-Yamamoto test in the next stage. In
12
order to obviate the possibility of spurious causality, TY causality analyses are run in
trivariate models. That is, causality between two series is test, conditional upon the presence
of a third one. The abovementioned theories of possible interactions between financial depth,
economic development and trade openness provides the ground for such specifications.
(a) Financial deepening – economic development causality
The theory suggests that financial depth may be either a critical factor or a negligible one for
economic development. The former supports for the supply-leading or bidirectional causality
hypothesis while the latter supports for demand-following or insignificant financial depth-
economic development causation. Table 4a presents the results of the interaction between
DEPTH and OUTPUT, conditional on OPENNESS. The results generally show no sign of
autocorrelation or multicollinearity and are statistically significant and stable, in particular
with respect to the lag orders chosen in accordance with the causality testing procedure.
The analysis reveals relatively strong causal linkages between financial depth and economic
development for the investigated countries. Particularly, the evidence of finance-led
economic development is found in the cases of Malaysia and New Zealand. For China,
Indonesia and Japan, the findings suggest a feedback relationship between financial
deepening and economic development, that is, bidirectional finance-growth causality. For
Australia, Nepal and Philippines, the results support the demand-following hypothesis, so
financial depth is caused by economic development. With respect to the other countries in the
sample, the analysis does not show any significant causal linkages between financial depth
and economic development.
[Please insert Table 4a here]
Based on the findings, it can be concluded that there are indeed interactions between financial
depth and economic development in Asia and Pacific countries, as theories on the finance-
growth nexus imply. The results fit in reasonably well, thanks to generally continuous
improvements in financial depth and related institutions in AP countries.
Even for the poorest economy in the region like Nepal, over the past 20 years the country’s
financial sector has become deeper and the number and type of financial intermediaries have
grown rapidly. The Nepalese financial system witnessed a large jump in terms of number of
13
financial institutions after financial liberalization that started in the mid of the 1980s. This has
translated into an increasing trend of financial development in this country. His Majesty’s
Government (HMG) of Nepal has recently implemented further financial reforms, improved
public expenditure management, strengthened anticorruption institutions, and improved
financial sector regulatory framework. Financial sector reform has been undertaken to
improve the performance of loss making public financial institutions. Other major recent
initiatives include (a) governance reform program, (b) decentralization and transfer of local
level activities in basic education, basic health, agriculture extension, drinking water, and
rural infrastructure to local bodies and the community level management committees (c)
strengthening of monitoring and evaluation system and establishment of poverty monitoring
mechanism and (d) privatisation of public enterprises.2 Recent reforms have made the
Nepalese banking sector more stable. Still, access to financial services remains limited for
many people in Nepal and thus HMG needs measures to strengthen further its reform process.
Overall, it appears reasonable to find that for the considered Asia-Pacific countries, financial
sectors interact with real sectors quite significantly. The findings suggest that a policy focus
on deepening financial sector to stimulate economic development seems to be justified.
(b) Financial deepening – trade openness causality
Theoretical considerations suggest that finance may unilaterally lead openness or that
openness may induce financial development. A nexus between finance and openness may
additionally allow for bidirectional causality. More sceptical views, however, may suggest no
evidence of significant causality between finance and openness. Table 4b shows the results
for causal inferences of DEPTH and OPENNESS, controlling for OUTPUT. The results
again show no sign of autocorrelation or multicollinearity and appear to be stable, particularly
with respect to the chosen lag orders.
The findings appear to confirm the existence of a nexus between financial depth and trade
openness. Nevertheless, this study is unable to identify a predominant causation pattern for
many investigated countries. Specifically, the evidence of the hypothesis that financial depth
Granger causes trade openness is found for India and Malaysia. Meanwhile, the findings
2 Refer to the progress report on Nepal available at: http://www.un.org/special-rep/ohrlls/ldc/MTR/Nepal.pdf for
further details.
14
suggest that trade openness has unilaterally influenced financial depth in the cases of China,
Indonesia, Japan, Korea, Nepal, New Zealand and Philippines. For the rest of the countries
included in the sample, the results do not indicate any stable long-run causality. This finding
may not be surprising. A possible explanation for the lack of causal linkage between trade
and financial depth (defined as formal finance) could be that, especially for low-income
countries, informal finance could be important for trade (see Roy et al., 2014).
[Please insert Table 4c here]
The findings thus offer support for theoretical and empirical considerations on financial
deepening – trade openness linkages. Policies that aim at enhancing a country’s financial
depth are thus likely to significantly shape trade structures as a by-product. Along the line of
this argument, policies that are targeted at increasing the levels of openness can be expected
to have substantial finance-promoting effects.
However, the effect of financial deepening – trade openness linkages on general economic
development in the investigated AP countries appears to be rather marginal. On the one hand,
the influence of trade openness on financial depth has not translated into economic
development, as shown by the previous results. Only in the cases of China, Indonesia, Japan
and New Zealand does it seem that trade openness has interacted with financial depth, which
in turn has contributed to economic development. In other words, there is rather limited
evidence of an indirect effect of trade openness on economic development via the channel of
financial development.
On the other hand, neither does this study find strong evidence of the hypothesis that finance-
induced advances in trade openness have translated into enhanced economic performance.
This is apparent from the causality analysis results of OUTPUT and OPENNESS, conditional
on DEPTH, which is presented in Table 4c. Here in most cases either trade openness
Granger-causes economic development or both series share a feedback relationship. When
combining the findings from Table 4b and 4c, the results indicate that in all cases, no indirect
effect of financial deepening on economic development through the channel of trade
openness can be demonstrated.
[Please insert Table 4b here]
15
(c) Robustness
This study relies on a composite indicator of financial depth. While the use of this index
yields some advantages as discussed, it may also have disadvantages. Such shortcomings
may, for example, be associated with a limited interpretability of the index. As such, this
study once again performs the empirical analysis using liquid liabilities to GDP (LLGDP) as
the indicator of financial depth. This measure is a more traditional finance indicator and has
been employed by a number of studies in literature (e.g. King and Levin, 1993). Using
LLGDP instead of DEPTH should help to assess the validity of the previous empirical
findings. The same econometric procedure as introduced is followed. In general, the
robustness findings confirm the previous results. Unit root and cointegration tests show
almost identical patterns when using LLGDP instead of DEPTH. Causal linkages between the
variables in the sample are also qualitatively the same.
(d) Discussion and Policy implications
The findings indicate (1) a relatively strong linkage between financial depth and economic
development, (2) a significant but somewhat weaker linkage between financial depth and
trade openness and (3) a sceptical linkage between trade openness and economic
development, for most of the sample.
The findings support the empirical studies that find strong linkages between financial depth
and economic development (e.g., King and Levin, 1993; Robinson, 1952; Berthelemy and
Varoudakis, 1996). Still, other studies do not find significant links (e.g., Chandavarkar,
1992). It might be concluded that the different findings of studies on financial deepening-
economic growth causality are attributable to different country samples rather than
differences in methodology. This is because the robustness check indicates that the findings
in this study are not random, so different methodologies are less likely to account for varying
results than different country samples. Generally, the findings of this study support the view
that ‘one size does not fit all’ when analysing financial deepening-economic development
interactions (Rioja and Valev, 2004). That is, the actual effect of financial depth on economic
development (and vice versa) seems to depend on the level of financial development. When
the level of financial development is low, the effect of finance on economic development is
uncertain (Rioja and Valev, 2004).
16
The analysis of this study suggests that (4) only few of the selected AP countries in the
sample have actually benefited directly or indirectly from financial development. Meanwhile,
(5) the direct and indirect effect of trade openness on economic development is somewhat
more pronounced. As a consequence, development strategies that unilaterally focus either
financial or trade sector development do not appear to be feasible for the countries in the
sample. Though the findings suggest that finance and finance-related policies have not
mattered significantly in the past, they do not imply that finance is irrelevant to development
in the future. This is because evidence from other parts of the world does reveal that financial
deepening promotes economic development. One possible explanation for the lack of causal
linkage between financial depth (defined as formal finance) and real sectors in this case could
be that, especially for the low-income countries, informal finance plays an important role.
Much like trade, financial development (or the state of the financial sector in a country) is
outcomes, in large parts, of policies such as financial reforms. In other words, obstacles to
economic development such as poor institutions or political instability are also obstacles to
the development of financial markets. As such, economic policies that aim at removing
growth obstacles may also be helpful in promoting financial development, thereby helping to
overcome financial system deficiencies and benefiting finance-growth dynamics. Possible
promising development strategies are greater political and macroeconomic stability or
improved institutional quality, all of which could in turn positively impact financial
development (e.g., Montiel, 2003; Demetriades and Law, 2006). Hence, a general approach
taking into account fundamental determinants of development seems to be more appropriate
for the sample of AP countries in this study. At the same time, these countries could gain
more from trade by implementing such policies.
5. CONCLUDING REMARKS
This study draws on conflicting considerations about the relationships between financial
deepening, economic development and trade openness by testing for causality for 14 AP
countries. Particularly, this study conducts a principal component analysis to obtain a broad
indicator of financial depth. The research employs unit root and cointegration tests to analyse
the properties of the investigated time series and to identify possible long-run relationships
between them. Toda-Yamamoto non-Granger causality test within unrestricted VAR
frameworks is then used due to its methodological advantages over standard causality tests.
17
The principal findings of the study are: (1) Financial depth, trade openness and economic
development do not share significant long-run relationships for the majority of the sample
(except for Korea, Japan, Nepal, China and Israel). (2) The results reveal a strong linkage
between financial depth and economic development. Yet, there is only some support for the
hypothesis of finance-led development. For most countries it is detected only a demand-
following or an insignificant relationship between financial depth and economic
development. (3) The connection between financial depth and trade openness is significant
but somewhat weaker. This study is unable to find enough evidence to suggest that either
financial deepening has promoted economic development indirectly via influencing trade
openness or that trade openness has contributed to economic development as a by-product of
its impact on financial development. (4) Finally, this study finds that there is a sceptical
linkage between trade openness and economic development, for most of the countries.
As a result, this research does not advocate development strategies that prioritize unilaterally
either financial or trade sector development. Instead, it supports a more balanced policy
approach that takes into account other fundamental development factors, for example
political or macroeconomic stability, or institutional quality. A general approach toward
strengthening of these factors may also help to reduce deficiencies in financial systems, so
countries in Asia and the Pacific may benefit from financial deepening in the future. Such an
approach should also help countries to gain more from trade openness.
Word Count: 5,616 words.
18
Table 1: Summary statistics and results of principal component analysis
Income category
Country
(data
availability)
DEPTH
(principal
component), %
Component matrix
DBMA LL PC
East Asia and Pacific
(Lower-middle-income
economies)
Indonesia
(1981-2011) 65.63 0.507 0.500 0.701
South Asia (Lower-
middle-income economies)
India (1961-
2011) 96.09 0.571 0.580 0.581
High-income OECD
members
Japan (1961-
2011) 60.28 -0.164 0.726 0.668
High-income OECD
members
Korea (1971-
2011) 89.75 0.571 0.585 0.576
East Asia and Pacific
(Upper-middle-income
economies)
Malaysia
(1961-2011) 71.10 0.349 0.667 0.658
East Asia and Pacific
(Lower-middle-income
economies)
Philippines
(1961-2011) 69.58 0.516 0.584 0.626
East Asia and Pacific
(Upper-middle-income
economies)
Thailand
(1966-2011) 91.51 0.561 0.585 0.586
East Asia and Pacific
(Upper-middle-income
economies)
China (1987-
2011) 92.36 0.559 0.587 0.586
High-income OECD
members
New Zealand
(1961-2010) 86.42 0.570 0.585 0.576
High-income OECD
members
Australia
(1961-2011) 81.37 0.524 0.587 0.617
South Asia (Lower-
middle-income economies)
Pakistan
(1961-2011) 76.62 0.524 0.596 0.609
High-income OECD Israel (1961- 89.11 0.540 0.588 0.602
19
members 2009)
South Asia (Lower-income
economies)
Nepal (1964-
2011) 67.98 0.207 0.690 0.694
South Asia (Lower-
middle-income economies)
Sri Lanka
(1961-2011) 82.20 0.565 0.573 0.593
Note: Data for the individual finance indicators is taken from the updated and expanded version of Financial
Development and Structure Database (FDSD). The column DEPTH contains the value of the initial eigenvalues
as a percentage of the total variance the first principal component contains (percentage of variance criterion) that
represents the composite indicator of financial depth. Following the standard income measurement of the World
Bank as taken from Beck et al. (2001), Indonesia, India, Philippines, Pakistan and Sri Lanka can be classified as
Lower Middle Income countries; Thailand, Malaysia and China can be classified as Upper Middle Income
countries; Japan, Korea, New Zealand, Australia and Israel can be classified as High Income countries; Nepal is
classified as Lower Income country.
20
Table 2: Phillips-Perron unit root test statistics
Country
Log level First log difference
INT INT&
TREND INT
INT&
TREND
Australia OUTPUT -0.861 -2.139 -6.004*** -6.022***
OPENNESS -0.201 -1.988 -5.954*** -5.885***
DEPTH -0.272 -1.619 -5.575*** -5.516***
China OUTPUT 1.031 -1.748 -2.567 -2.736
OPENNESS 0.064 -1.728 -4.673*** -4.596***
DEPTH -1.133 -1.319 -3.335** -3.378*
India OUTPUT 9.462 2.090 -5.834*** -8.231***
OPENNESS 0.462 -2.142 -6.065*** -6.151***
DEPTH -0.425 -1.781 -5.177*** -5.126***
Indonesia OUTPUT -0.710 -1.738 -4.051*** -3.968**
OPENNESS -0.235 -1.890 -4.988*** -4.988***
DEPTH -2.594 -2.211 -2.967** -3.024
Israel OUTPUT -2.675 -2.512 -5.207*** -5.349***
OPENNESS -1.043 -1.612 -7.268*** -7.294***
DEPTH -2.041 -3.242* -6.248*** -6.214***
Japan OUTPUT -6.449 -2.223 -3.910*** -5.344***
OPENNESS -1.414 -1.219 -6.547*** -6.655***
DEPTH -3.440** -2.227 -4.357*** -4.243***
Korea OUTPUT -1.694 -0.603 -5.320*** -5.563***
OPENNESS -3.620*** -3.537** -4.751*** -5.148***
DEPTH -0.344 -3.099 -5.005*** -4.859***
Malaysia OUTPUT -0.943 -1.931 -6.161*** -6.162***
21
OPENNESS -0.269 -2.339 -5.645*** -5.578***
DEPTH -0.973 -2.036 -7.147*** -7.656***
Nepal OUTPUT 2.181 -1.300 -8.756*** -10.318***
OPENNESS 0.449 -2.772 -7.386*** -7.476***
DEPTH -1.017 -2.981 -6.346*** -6.306***
New Zealand OUTPUT -1.257 -2.970 -5.198*** -5.090***
OPENNESS -1.401 -1.324 -7.010*** -7.437***
DEPTH -1.127 -2.324 -6.112*** -6.135***
Philippines OUTPUT -0.742 -1.648 -3.826*** -3.778**
OPENNESS -0.632 -2.246 -6.263*** -6.182***
DEPTH -1.871 -2.472 -4.612*** -4.548***
Pakistan OUTPUT -0.742 -1.648 -3.826*** -3.778***
OPENNESS -0.705 -2.418 -5.534*** -5.496***
DEPTH -3.544** -2.952 -3.906*** -4.166***
Sri Lanka OUTPUT 3.500** -1.070 -4.945*** -6.131***
OPENNESS -0.322 -2.965 -6.451*** -6.411***
DEPTH -1.385 -2.492 -4.669*** -4.643***
Thailand OUTPUT -0.931 -1.562 -4.389*** -4.422***
OPENNESS -0.154 -2.533 -5.513*** -5.460***
DEPTH -1.054 -1.218 -3.966*** -3.853**
Note: ***, ** and * denote significance at 1%, 5% and 10% levels, respectively. The critical values are taken
from MacKinnon (1996).
22
Table 3a: Gregory-Hansen cointegration test: Dependent variable DEPTH
Level shift
C
Level shift with
trend
C/T
Regime shift
C/S
Australia ADF* -3.599 (2) -3.941 (2) -5.256 (3)*
[1971] [1972] [1985]
-17.497 -26.137 -28.131
[1973] [1976] [1984]
-3.216 -4.065 -4.191
[1973] [1976] [1985]
China ADF* -4.347 (0) -4.620 (0) -5.794 (3)**
[2005] [2005] [2000]
-25.229 -26.483 -26.000
[2006] [2006] [2005]
-4.893** -5.342** -5.097
[2006] [2006] [2005]
India ADF* -3.685 (2) -3.906 (2) -3.605 (1)
[1972] [1984] [1973]
-19.974 -19.138 -19.810
[1968] [1986] [1973]
-3.363 -3.345 -3.328
[1968] [1986] [1982]
Indonesia ADF* -4.101 (0) -6.039 (1)*** -4.099 (0)
[1999] [1993] [1997]
-20.826 -21.689 -21.705
[1999] [1993] [1997]
-4.176 -4.468 -4.174
[1999] [1993] [1997]
Israel ADF* -4.996 (6)** -5.001 (2) -5.224 (6)
[1999] [1998] [1999]
-27.472 -26.895 -27.619
23
[1998] [1998] [1998]
-4.370 -4.422 -4.431
[1998] [1981] [1981]
Japan ADF* -5.389 (1)** -5.662 (1)** -5.373 (1)*
[2002] [2002] [2002]
-31.996 -34.695 -31.974
[2002] [2002] [2002]
-4.632 -5.034* -4.629
[2002] [2002] [2002]
Korea ADF* -5.641 (3)*** -6.015 (0)*** -5.610 (0)**
[1983] [1978] [1979]
-34.835 -39.408 -37.406
[1979] [1979] [1979]
-5.545*** -6.233*** -5.716**
[1979] [1979] [1979]
Malaysia ADF* -4.747 (1)* -4.832 (1) -4.994 (1)
[1992] [1969] [1992]
-26.545 -30.979 -27.965
[1993] [1968] [1986]
-4.133 -4.529 -4.320
[1994] [1968] [1986]
Nepal ADF* -5.996 (4)*** -4.729 (4) -5.572 (1)**
[1971] [1982] [1969]
-34.875 -24.547 -35.281
[1967] [1984] [1969]
-5.161** -4.097 -5.482*
[1967] [1982] [1974]
New Zealand ADF* -4.538 (3) -5.063 (3)* -4.246 (3)
[1988] [1988] [1992]
-17.670 -29.481 -23.926
[1975] [1999] [1975]
24
-3.253 -4.411 -3.995
[1975] [1999] [1972]
Pakistan ADF* -5.412 (1)*** -5.835 (1)*** -4.547 (1)
[1982] [1967] [1987]
-22.867 -20.992 -23.135
[1983] [1979] [1973]
-3.665 -3.553 -3.664
[1983] [1968] [1973]
Philippines ADF* -4.973 (1)** -5.415 (1)** -4.902 (1)
[1983] [1993] [1985]
-23.928 -26.350 -30.562
[1984] [1984] [1984]
-4.171 -4.707 -4.674
[1984] [1994] [1984]
Sri Lanka ADF* -4.375 (1) -5.238 (1)* -5.377 (1)*
[1968] [1997] [1997]
-18.155 -25.903 -27.126
[1967] [1995] [1995]
-3.125 -3.932 -4.138
[1995] [1995] [1995]
Thailand ADF* -5.261 (1)** -5.218 (1)* -6.080 (1)***
[2002] [2002] [2002]
-17.758 -19.988 -24.722
[2003] [1984] [1986]
-3.224 -3.393 -4.005
[2003] [1984] [1986]
25
Table 3b: Gregory-Hansen cointegration test: Dependent variable OUTPUT
Level shift
C
Level shift with trend
C/T
Regime shift
C/S
Australia ADF* -3.742 (2) -3.860 (2) -4.052 (2)
[1984] [1986] [1984]
-12.133 -18.926 -17.370
[1970] [1968] [1986]
-2.567 -3.378 -3.178
[1996] [1968] [1986]
China ADF* -3.832 (2) -4.654 (2) -4.793 (0)
[2006] [2006] [2006]
-26.603 -14.670 -26.471
[2006] [2006] [2003]
-5.247** -3.023 -5.225
[2006] [2006] [2003]
India ADF* -3.134 (0) -4.301 (0) -3.985 (0)
[1975] [1975] [1986]
-16.498 -28.628 -26.377
[1975] [1975] [1986]
-3.105 -4.345 -4.052
[1975] [1975] [1986]
Indonesia ADF* -4.816 (3)* -5.519 (1)** -5.480 (3)*
[1998] [1994] [1992]
-22.803 -22.858 -24.974
[1985] [1993] [1992]
-4.169 -4.426 -4.560
[1985] [1993] [1992]
Israel ADF* -4.522 (4) -4.886 (0) -5.531 (0)**
[1978] [1969] [1974]
-22.498 -30.918 -38.189
[1976] [1969] [1974]
26
-4.004 -4.947 -5.584**
[1976] [1969] [1974]
Japan ADF* -4.607 (4) -4.503 (4) -5.376 (4)*
[2001] [2001] [1990]
-19.602 -21.123 -28.637
[2002] [1967] [1972]
-3.315 -3.458 -4.288
[1975] [1967] [1972]
Korea ADF* -5.125 (0)** -3.626 (1) -5.303 (1)*
[1979] [1988] [1984]
-33.637 -20.344 -33.064
[1979] [1988] [1984]
-5.239** -3.392 -5.020
[1979] [1988] [1984]
Malaysia ADF* -4.578 (0) -3.850 (0) -4.956 (1)
[1996] [1967] [1995]
-22.946 -25.658 -24.397
[1995] [1967] [1980]
-4.625 -4.041 -4.800
[1996] [1967] [1995]
Nepal ADF* -4.640 (1) -4.503 (0) -7.164 (0)***
[1968] [1969] [1985]
-27.437 -29.480 -49.709
[1968] [1968] [1985]
-4.364 -4.552 -7.241***
[1969] [1969] [1985]
New Zealand ADF* -5.215 (3)** -5.388 (3)** -5.390 (3)*
[2001] [1996] [1998]
-21.308 -28.871 -24.154
[2002] [1999] [2000]
-3.455 -4.402 -3.809
27
[2002] [1999] [2000]
Pakistan ADF* -3.343 (1) -4.222 (1) -3.159 (1)
[1982] [1987] [1987]
-16.853 -23.122 -17.674
[1983] [1986] [1983]
-2.751 -3.771 -3.198
[1983] [1986] [1983]
Philippines ADF* -3.019 (6) -3.682 (3) -2.563 (6)
[1980] [1989] [1968]
-8.454 -17.941 -8.348
[1987] [1985] [1991]
-1.902 -3.265 -1.705
[1985] [1985] [2002]
Sri Lanka ADF* -4.169 (6) -3.519 (3) -3.536 (2)
[1998] [1999] [1995]
-14.299 -16.488 -18.649
[2002] [2002] [1991]
-2.759 -2.705 -3.217
[2002] [2002] [1994]
Thailand ADF* -5.163 (1)** -4.719 (3) -5.673 (3)**
[1974] [1989] [1989]
-22.677 -20.641 -26.723
[1975] [1990] [1979]
-3.888 -3.590 -4.696
[1975] [1990] [1979]
28
Table 3c: Gregory-Hansen cointegration test: Dependent variable OPENNESS
Level shift
C
Level shift with trend
C/T
Regime shift
C/S
Australia ADF* -4.372 (5) -4.619 (3) -4.054 (3)
[1980] [1995] [1983]
-20.756 -24.987 -18.459
[1975] [1995] [1975]
-3.359 -3.901 -3.079
[1975] [1994] [1974]
China ADF* -4.608 (0) -5.732 (0)** -5.786 (0)**
[2006] [2003] [2004]
-24.101 -29.403 -32.611
[2006] [2003] [2003]
-4.712 -5.856*** -6.702***
[2006] [2003] [2003]
India ADF* -4.265 (6) -4.692 (3) -4.150 (3)
[1973] [1979] [1974]
-20.448 -24.784 -20.519
[1976] [1976] [1976]
-3.448 -3.962 -3.503
[1975] [1975] [1975]
Indonesia ADF* -4.890 (3)* -7.617 (3)*** -5.391 (3)*
[1998] [2004] [1996]
-23.123 -23.096 -20.891
[1985] [1985] [1987]
-4.117 -4.113 -3.827
[1985] [1985] [1987]
Israel ADF* -5.176 (0)** -5.598 (0)** -5.787 (0)**
[1976] [1974] [1974]
-34.351 -39.376 -41.338
[1975] [1974] [1974]
29
-5.484*** -5.734** -5.939**
[1975] [1974] [1974]
Japan ADF* -4.016 (6) -4.774 (0) -4.382 (0)
[1968] [1973] [1973]
-21.564 -31.798 -29.326
[1975] [1973] [1973]
-3.566 -4.802 -4.382
[1975] [1973] [1973]
Korea ADF* -4.774 (0)* -5.973 (1)*** -5.387 (1)*
[1978] [1978] [1984]
-31.528 -39.230 -32.274
[1978] [1978] [1984]
-4.858* -6.061*** -4.825
[1978] [1978] [1984]
Malaysia ADF* -5.024 (6)** -4.787 (1) -5.111 (6)
[1980] [1967] [1980]
-22.198 -31.466 -28.662
[1967] [1967] [1975]
-4.164 -4.759 -4.308
[1975] [1967] [1975]
Nepal ADF* -4.348 (2) -4.171 (0) -4.329 (2)
[1978] [1967] [1978]
-25.120 -28.062 -27.839
[1976] [1967] [1976]
-3.992 -4.191 -4.253
[1976] [1967] [1973]
New Zealand ADF* -4.400 (3) -4.637 (1) -4.314 (1)
[1975] [1976] [1977]
-24.744 -32.420 -27.796
[1976] [1999] [1977]
-3.913 -4.821 -4.192
30
[1976] [1999] [1977]
Pakistan ADF* -4.047 (1) -5.274 (5)* -4.697 (0)
[1982] [1978] [1973]
-21.582 -21.478 -30.615
[1982] [1977] [1973]
-3.443 -3.465 -4.688
[1982] [1977] [1973]
Philippines ADF* -5.579 (1)*** -6.439 (1)*** -5.279 (1)*
[1983] [1976] [1983]
-27.139 -28.881 -32.936
[1983] [1975] [1989]
-4.376 -4.829 -4.590
[1984] [1975] [1990]
Sri Lanka ADF* -3.826 (5) -4.019 (5) -4.447 (1)
[1980] [1980] [1976]
-17.516 -19.993 -28.747
[1976] [1967] [1976]
-3.520 -3.580 -4.252
[1976] [1967] [1976]
Thailand ADF* -4.188 (1) -4.387 (1) -5.367 (0)*
[1976] [1976] [1973]
-23.227 -24.206 -37.111
[1975] [1975] [1973]
-3.817 -3.942 -5.428*
[1975] [1975] [1973]
Note: VAR consists of DEPTH, OPENNESS and OUTPUT (m=2).*, ** and *** denote significance, i.e.
rejection of the null hypothesis of no cointegration at 10%, 5% and 1% levels, respectively. Numbers in (.) are
lag orders to include in equations. Lag lengths are determined automatically based on AIC (max=6). Time
breaks are in [.]
Note: Approximate asymptotic critical values for C, C/T and C/S respectively: m=2: -5.44, -5.80, -5.97 for
ADF* and and -57.01, -64.77, -68.21 for (at 1% level); -4.92, -5.29, -5.50 for ADF* and and -46.98, -
53.92, -58.33 for (at 5% level); -4.69, -5.03, -5.23 for ADF* and and -42.49, -48.94, -52.85 for (at
10% level). Critical values are taken from Table 1, page 109, Gregory and Hansen, 1996, Residual-based tests
for cointegration in models with regime shifts, Journal of Econometrics, 70, p. 99-126.
31
Table 4a: Toda-Yamamoto non-Granger causality test: DEPTH AND OUTPUT
Null hypothesis Lag Wald statistic p-value
Australia DEPTH OUTPUT 2 1.249 0.536
OUTPUT DEPTH 2 5.125* 0.077
China DEPTH OUTPUT 2 8.047** 0.018
OUTPUT DEPTH 2 5.154* 0.076
India DEPTH OUTPUT 4 1.828 0.767
OUTPUT DEPTH 4 5.329 0.255
Indonesia DEPTH OUTPUT 5 9.498* 0.091
OUTPUT DEPTH 5 11.638** 0.040
Israel DEPTH OUTPUT 3 3.314 0.346
OUTPUT DEPTH 3 5.451 0.142
Japan DEPTH OUTPUT 2 9.072** 0.011
OUTPUT DEPTH 2 6.999** 0.030
Korea DEPTH OUTPUT 1 0.158 0.691
OUTPUT DEPTH 1 0.581 0.446
Malaysia DEPTH OUTPUT 3 8.787** 0.032
OUTPUT DEPTH 3 0.880 0.830
Nepal DEPTH OUTPUT 2 1.736 0.420
OUTPUT DEPTH 2 4.638* 0.098
New Zealand DEPTH OUTPUT 2 5.651* 0.059
OUTPUT DEPTH 2 1.532 0.465
Philippines DEPTH OUTPUT 2 0.378 0.828
OUTPUT DEPTH 2 8.646** 0.013
Pakistan DEPTH OUTPUT 2 2.591 0.274
OUTPUT DEPTH 2 0.821 0.663
Sri Lanka DEPTH OUTPUT 2 1.096 0.578
OUTPUT DEPTH 2 2.569 0.277
Thailand DEPTH OUTPUT 2 0.350 0.839
OUTPUT DEPTH 2 0.110 0.946
Note: VAR consists of DEPTH, OUTPUT and OPENNESS (satisfy stability condition). The maximum order of
integration among the variables of interest is 1.Lag lengths are determined based on Schwarz Information
Criterion (SIC). *, ** and *** denote significance, i.e. rejection of the null hypothesis of no causality at 10%,
5% and 1% levels, respectively.
32
Table 4b: Toda-Yamamoto non-Granger causality test: DEPTH AND OPENNESS
Null hypothesis Lag Wald statistic p-value
Australia DEPTH OPENNESS 2 2.585 0.275
OPENNESS DEPTH 2 1.020 0.600
China DEPTH OPENNESS 2 0.935 0.627
OPENNESS DEPTH 2 7.497** 0.024
India DEPTH OPENNESS 4 13.788*** 0.008
OPENNESS DEPTH 4 2.304 0.680
Indonesia DEPTH OPENNESS 5 4.057 0.541
OPENNESS DEPTH 5 12.549** 0.028
Israel DEPTH OPENNESS 3 1.761 0.623
OPENNESS DEPTH 3 1.047 0.790
Japan DEPTH OPENNESS 2 0.609 0.737
OPENNESS DEPTH 2 5.038* 0.080
Korea DEPTH OPENNESS 1 0.403 0.526
OPENNESS DEPTH 1 4.205** 0.040
Malaysia DEPTH OPENNESS 3 10.120** 0.018
OPENNESS DEPTH 3 0.280 0.960
Nepal DEPTH OPENNESS 2 2.957 0.228
OPENNESS DEPTH 2 5.249* 0.073
New Zealand DEPTH OPENNESS 2 0.933 0.627
OPENNESS DEPTH 2 4.842* 0.089
Philippines DEPTH OPENNESS 2 0.213 0.899
OPENNESS DEPTH 2 5.011* 0.082
Pakistan DEPTH OPENNESS 2 0.679 0.712
OPENNESS DEPTH 2 0.743 0.690
Sri Lanka DEPTH OPENNESS 2 1.898 0.387
OPENNESS DEPTH 2 1.617 0.445
Thailand DEPTH OPENNESS 2 1.164 0.559
OPENNESS DEPTH 2 0.226 0.893
Note: VAR consists of DEPTH, OUTPUT and OPENNESS (satisfy stability condition). The maximum order of
integration among the variables of interest is 1.Lag lengths are determined based on Schwarz Information
Criterion (SIC). *, ** and *** denote significance, i.e. rejection of the null hypothesis of no causality at 10%,
5% and 1% levels, respectively.
33
Table 4c: Toda-Yamamoto non-Granger causality test: OPENNESS AND OUTPUT
Null hypothesis Lag Wald statistic p-value
Australia OPENNESS OUTPUT 2 3.778 0.151
OUTPUT OPENNESS 2 2.415 0.299
China OPENNESS OUTPUT 2 5.335* 0.069
OUTPUT OPENNESS 2 1.237 0.539
India OPENNESS OUTPUT 4 1.498 0.827
OUTPUT OPENNESS 4 2.310 0.679
Indonesia OPENNESS OUTPUT 5 9.315* 0.097
OUTPUT OPENNESS 5 13.077** 0.023
Israel OPENNESS OUTPUT 3 7.755* 0.051
OUTPUT OPENNESS 3 2.721 0.437
Japan OPENNESS OUTPUT 2 1.835 0.400
OUTPUT OPENNESS 2 0.779 0.677
Korea OPENNESS OUTPUT 1 2.846 0.092
OUTPUT OPENNESS 1 0.035 0.853
Malaysia OPENNESS OUTPUT 3 0.305 0.959
OUTPUT OPENNESS 3 5.928 0.115
Nepal OPENNESS OUTPUT 2 10.567*** 0.005
OUTPUT OPENNESS 2 1.451 0.484
New Zealand OPENNESS OUTPUT 2 1.450 0.484
OUTPUT OPENNESS 2 3.804 0.149
Philippines OPENNESS OUTPUT 2 4.667* 0.097
OUTPUT OPENNESS 2 3.371 0.185
Pakistan OPENNESS OUTPUT 2 3.165 0.205
OUTPUT OPENNESS 2 1.561 0.458
Sri Lanka OPENNESS OUTPUT 2 0.447 0.800
OUTPUT OPENNESS 2 1.266 0.531
Thailand OPENNESS OUTPUT 2 0.649 0.723
OUTPUT OPENNESS 2 2.161 0.339
Note: VAR consists of DEPTH, OUTPUT and OPENNESS (satisfy stability condition). The maximum order of
integration among the variables of interest is 1.Lag lengths are determined based on Schwarz Information
Criterion (SIC). *, ** and *** denote significance, i.e. rejection of the null hypothesis of no causality at 10%,
5% and 1% levels, respectively.
34
REFERENCES
Ang, J. B., & McKibbin, W. J. (2007). Financial liberalization, financial sector development
and growth: Evidence from Malaysia. Journal of Development Economics, 84, 215–233.
Beck, T. (2002). Financial development and international trade: Is there a link? Journal of
International Economics, 2, 107-131.
Bencivenga, V. R., & Smith, B. D. (1998). Economic development and financial depth in a
model with costly financial intermediation. Research in Economics, 52, 363-386.
Berthelemy, J. C., & Varoudakis, A. (1996). Economic growth, convergence clubs, and the
role of financial development. Oxford Economic Papers, 48, 300–328.
Blackburn, K., & Hung, V. T. Y. (1993). A Theory of Growth, Financial Development and
Trade. Economica, 65(257), 107.125.
Calderon, C., & Lin, L. (2003). The direction of causality between financial development and
economic growth. Journal of Development Economics, 72, 321–334.
Casselli, F., Esquivel, G., & Lefort, F. (1996). Reopening the convergence debate: a new look
at cross-country growth empirics. Journal of Economic Growth, 1(3), 363-89.
Chandavarkar, A. (1992). Of finance and development: Neglected and unsettled questions.
World Development, 20(1), 133–142.
Demetriades, P. O., & Law, S. H. (2006). Finance, institutions and economic development.
International Journal of Finance and Economics, 11, 245–260.
Do, Q. T., & Levchenko, A. A., (2004). Trade and financial development. World Bank
Working Paper 3347.
Edwards, S. (1998). Openness, productivity and growth: What do we really know? The
Economic Journal, 108, 383–398.
Engle, R. F., & Granger, C. W. J. (1987). Cointegration and error correction: Representation,
estimation and testing. Econometrica, 55(2), 251-76.
Giles, D. (1997). Causality between the measured and underground economies in New
Zealand. Applied Economics Letters, 4(1), 63-67.
35
Ghirmay, T. (2004). Financial development and economic growth in sub-Saharan African
countries: evidence from time series analysis. African Development Review, 16, 415–432.
Granger, C. W. J. (1969). Investigating causal relations by econometric models and cross-
spectral methods. Econometrica, 37, 424-438.
Gregory, A. W., & Hansen, B. E. (1996). Residual-based tests for cointegration in models
with regime shifts. Journal of Econometrics, 70, 99-126.
Gregory, A. W., Nason, J. M., & Watt, D. (1996). Testing for structural breaks in
cointegrated relationships. Journal of Econometrics, 71, 321-341.
Gries, T., Kraft, M., & Meierrieks, D., (2009). Linkages between financial deepening, trade
openness, and economic development: causality evidence from Sub-Saharan Africa. World
Development, 37(12), pp. 1849-1860.
Gujarati, D. N. (1995). Basic Econometrics, 3rd
edition. McGraw-Hill, Inc.
Hakkio, C. S., & Rush, M. (1991). Cointegration: How short is the long run? Journal of
International Money and Finance, 10, 571–581.
Harrison, A. (1996). Openness and growth: A time-series, cross-country analysis for
developing countries. Journal of Development Economics, 48, 419-447.
Johansen, S. (1988). Statistical analysis of cointegration vectors. Journal of Economic
Dynamics and Control, 12(2-3), 231-4.
Johansen, S., & Juselius, K. (1990). Maximum likelihood estimation and inference on
cointegration: With application to the demand for money. Oxford Bulletin of Economics
and Statistics, 52, 169-210.
King, R., & Levine, R. (1993). Finance and growth: Schumpeter might be right. Quarterly
Journal of Economics, 108(3), 717–737.
Kletzer, K., & Bardhan, P. K. (1987). Credit markets and patterns of international trade.
Journal of Development Economics, 27(1-2), 57-70.
Krugman, P. R. (1979). Increasing returns, monopolistic competition and international trade.
Journal of International Economics, 9(4), 469-479.
36
Krugman, P. R. (1980). Scale economies, product differentiation, and the pattern of trade.
American Economic Review, 70(5), 950-959.
La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2002). Government ownership of banks.
Journal of Finance, 57(1), 265–301.
Lau, E., & Baharumshah, A. Z. (2003). Sustainability of external imbalances: The case of
Malaysia. Singapore Economic Review, 48, 61-80.
Levine, R. (2005). Chapter 12: Finance and growth: theory, evidence and mechanisms. In
Phillipe Aghion and Steven Durlauf (Ed), Handbook of Economic Growth (pp. 865-934).
Amsterdam: North- Holland: Elsevier Publishers.
Loizides, J., & Vamvoukas, G. (2005). Government expenditure and economic growth:
evidence from tri-variate causality testing. Journal of Applied Economics, 8, 125–152.
Mavrotas, G., & Kelly, R. (2001). Old wine in new bottles: Testing causality between savings
and growth. The Manchester School, 69, 97-105.
Montiel, P. J. (2003). Development of financial markets and macroeconomic policy. Journal
of African Economies, 12(AERC Suppl. 2), 12–52.
Odhiambo, N. M. (2008). Financial depth, savings and economic growth in Kenya: A
dynamic causal linkage. Economic Modelling, 25, 704-713.
Pagano, M. (1993). Financial markets and growth: An overview. European Economic
Review, 37, 613–622.
Phillips, P., & Perron, P. (1988). Testing for a unit root in time series regression. Biometrika,
75(2); 335–346.
Rioja, F., & Valev, N. (2004). Does one size fit all? A reexamination of the finance and
growth relationship. Journal of Development Economics, 74(2), 429–447.
Robinson, J. (1952). The rate of interest, and other essays. London: Macmillan.
Romer, P.M. (1990). Endogenous Technological Change. Journal of Political Economy, 98,
S71-S101.
37
Roy, D., A. Munasib & Chen, X. (2014). Social Values and International Trade: The
Interplay between Social Values and Formal Finance. Weltwirtschaftliches Archiv.
Schumpeter, J. A. (1934). The Theory of Economic Development. Cambridge, MA:
Harvard University Press. (New York: Oxford University Press, 1961.) First published in
German, 1912.
Schumpeter, J. A. (1942). Capitalism, Socialism, and Democracy. New York: Harper and
Brothers. (Harper Colophon edition, 1976.)
Shahbaz, M. (2012). Does Trade Openness Affect Long Run Growth? Cointegration,
Causality and Forecast Error Variance Decomposition Tests for Pakistan. Economic
Modeling, 29(6), pp.2325-2339.
Toda, H. Y., & Yamamoto, T. (1995). Statistical inference in Vector Autoregressions with
possibly integrated processes. Journal of Econometrics, 66, 225-250.
Yanikkaya, H. (2003). Trade Openness and Economic Growth: A Cross-country Empirical
Investigation. Journal of Development Economics, 72(1), 57-89